Tuesday, March 25, 2008

Tiger users still a majority

Source: FlickrI was more than a bit surprised yesterday when I first saw the statistics published by Omnigroup regarding their userbase and what kind of operationg system their users are using. Sure enough, this data is not a definite indicator and probably only Apple can give more accurate data, but its probably good enough that we can at least draw some conclusions from it.

Omnigroup develops whole bunch of different applications that are focused both on business users and on home users. I haven't been able to find any statistics about how large their userbase actualy is and will for the time being presume that their large array of products gives a rather good sample of users - if, however, anyone finds this jumping to conclusions too quick, feel free to enlighten me in the comments. ;-)

On the first sight of data I was almost a bit disappointed about the number of Leopard users. 32% of Leopard users seemed rather low for a system that was relesed at the end of October and had brought with it a number of interesting and exciting features. But then I gave the whole idea of 32% another thought - if one third of users on a particular platform upgraded in roughly six months to the flagship version of some companys operating system, that doesn't sound that bad. Since Microsoft released Windows Vista in a recent past it could even make a rather interesting comparison of the speed of adoption of a new platform.

It turns out that only 8.7% of Windows users had opened their hearts to Vista (either voluntarily or when they purchased their new computers). Since Vista was available in retail stores on 30. January of 2007 this means that it was present on a market for about 420 days, which in turn means that its approximate rate of adoption is about .02% of Windows users per day or about .6% per month. For Leopard, this rate is almost ten times higher - since its introduction on 30th October of 2007 it was present on a market for about 150 days and was adopted at an avarage rate of .2 %. per day or 6% per month.

Based on the data that I used this means that Mac users are migrating to newest platform at almost ten times the rate of Windows users. This puts Apple in rather favourable position compared to Microsoft. If Microsoft has to maintain compatibility for such a long time (if adoption rate would be constant this would mean Windows XP would still be in usage for about 10+ years after Vista was released - until 2017 or more) it means it can not implament new features at a pace Apple can - or at least it can not make them requirements at until a large majority of its users migrate to a new platform. Apple on the other hand can implament new features and declare the old ones as deprecated far more easily and more often.

(Luckily Microsoft doesn't wait for 10+ years for its users to upgrade - Windows XP will enjoy full support until 14th April of 2009, after which it will enter Extended Support that will last until 2014, although personally I think Microsoft will extend both full support and Extended Support.)

For developers this means that when Apple said Carbon would not enjoy 64-bit benefits that can be found in Leopard and encouraged developers to get their hands dirty with Cocoa, developers can sleep a bit easier knowing that their users will shortly follow the development cycle that Apple dictates. On the other hand developers that had huge code bases written with Carbon probably enjoyed a number of sleepless nights after first hearing Steve Jobs about implamenting 64-bit support only for Cocoa.

At the end it is only fair to write a honest disclaimer - both Omnigroup and W3Schools say that their data is not to be fully trusted, to which I can only add that you should trust my conclusions even less. I only did some quick math with basic data that both of these companies provide online in an effort to see how users of different platforms view flagship products of their beloved company and how this love is translated into purchase of new operating system or new computer. This data clearly misses all those users that had migrated either to OS X or to Windows from other platforms (and probably a bunch of other edge cases). Basically my analysis is just a quick look on a subject that is almost impossible to analyse with great certainty and should be treated in that manner.